Removed ReadMe.md and Added Explanation inside all Notebooks

Author: iSiddharth20

notebooks/deep_learning/CNNImageClassification/1. PreProcessing.ipynb

@@ -5,7 +5,45 @@
    "id": "6a9fb5f2",
    "metadata": {},
    "source": [
-    "More At : https://github.com/iSiddharth20/DeepLearning-ImageClassification-Toolkit"
+    "## Welcome to the AI-ML-Jupyter-Notebooks repository! \n",
+    "### This guide will help you navigate and learn CNN Image Classification using TensorFlow and OpenCV.\n",
+    "\n",
+    "----\n",
+    "\n",
+    "#### Make sure to install necessary dependencies by running this command :\n",
+    "pip install -r requirements.txt\n",
+    "\n",
+    "---\n",
+    "\n",
+    "#### TensorFlow is Required to run this codebase.\n",
+    "Install using the Official Guige [HERE](https://www.tensorflow.org/install/pip).\n",
+    "\n",
+    "---"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "e9736387-67a9-454a-8a07-402e0f20b9ac",
+   "metadata": {},
+   "source": [
+    "### Understanding PreProcessing\n",
+    "- Load Dataset from 'DATA_DIR' Directory\n",
+    "- Creates DataFrame containing Full Paths of Images and their Class Labels\n",
+    "- (Change as per Requirement) Rescale Images to Computationally Efficient Resolution\n",
+    "- (Optional but Recommended) Extracts Largest Object from Image using 'image_processing' Function\n",
+    "    - Leverages Parallel Processing for Faster Results\n",
+    "- Compares Original and Rescaled+Processed Image SIde-By-Side to make necessary changes\n",
+    "- Converts Processed Images to NumPy Array and Exports as Pickle File\n",
+    "    - Verifies If Exported Pickle File is Appropriate through 10 Random Samples\n",
+    "- (Optional) Merge Certain Class Lables Together\n",
+    "- Split Data for Training, Testing, Validation with Stratify to ensure data balancing\n",
+    "    - Verify if Split is Appropriate through 2 random samples\n",
+    "- (Optional) Perform Random Oversampling on Data to reduce Biasness\n",
+    "    - Verify if Oversampling is Appropriate through 2 random samples\n",
+    "- Perform One-Hot-Encoding of Class Labels\n",
+    "- Training, Testing, Validation Data and One-Hot-Encoding is Exported as Pickle Files\n",
+    "\n",
+    "---"
    ]
   },
   {
@@ -569,10 +607,18 @@
     "    print('Export Unsuccessful for : OHE.pkl ') "
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "3fab41a4-8ee1-4d28-aa5e-55d6596c0c4a",
+   "metadata": {},
+   "source": [
+    "More At : https://github.com/iSiddharth20/DeepLearning-ImageClassification-Toolkit"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "0e1e3bfb-f1f8-4db6-8943-da8fce55b769",
+   "id": "27bc0e74-0129-4052-b8c1-5649b0b500bd",
    "metadata": {},
    "outputs": [],
    "source": []

notebooks/deep_learning/CNNImageClassification/2. Training_CNN_ResNet50.ipynb

@@ -2,10 +2,23 @@
  "cells": [
   {
    "cell_type": "markdown",
-   "id": "bf037fe5",
+   "id": "dd400876-f589-4f9e-8a48-326b58f26389",
    "metadata": {},
    "source": [
-    "More At : https://github.com/iSiddharth20/DeepLearning-ImageClassification-Toolkit"
+    "### Understanding Model Creating and Training of CNN ResNet50\n",
+    "- Training, Testing, Validation Data and One-Hot-Encoding are Imported\n",
+    "    - All Data is converted to TensorFlow Format\n",
+    "- Learning Rate Scheduler is Defined (Change If Desired) \n",
+    "- Stochastic Gradient Descent with Momentum is Used as Optimizer  (Change If Desired) \n",
+    "- Added Data Augmentation Techniques to improve Model Learning  (Change If Desired) \n",
+    "- Base Model (CNN ResNet50) is Loded from TensorFlow Library\n",
+    "    - Custom Optimal Changes have been made to the Structure\n",
+    "    - Final Model is Compiled\n",
+    "- Final Model is Trained\n",
+    "    - Final Model with Lowest Validation Loss is Exported as a '.h5' file\n",
+    "- Traning Time (In Seconds) is Displayed\n",
+    "\n",
+    "---"
    ]
   },
   {
@@ -298,10 +311,18 @@
     "print('Testing : ', model.evaluate(test_dataset))"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "f7082894-40c3-49a9-85df-fe6221c5079c",
+   "metadata": {},
+   "source": [
+    "More At : https://github.com/iSiddharth20/DeepLearning-ImageClassification-Toolkit"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "abde9303-25fa-4ea5-84d7-c946da47a59d",
+   "id": "86eef81e-0ff9-4b6d-b6c0-b955c49b9a15",
    "metadata": {},
    "outputs": [],
    "source": []

notebooks/deep_learning/CNNImageClassification/3. Verification and Confusion Matrix.ipynb

@@ -2,10 +2,17 @@
  "cells": [
   {
    "cell_type": "markdown",
-   "id": "9fc5ca60",
+   "id": "1fb9033f-927a-4954-acc6-d38357bc3fd0",
    "metadata": {},
    "source": [
-    "More At : https://github.com/iSiddharth20/DeepLearning-ImageClassification-Toolkit"
+    "### Understanding Verification and Confusion Matrix\n",
+    "- Trained Model and One-Hot-Encoding are Imported\n",
+    "- Entire Dataset is Run through the Trained Model to get Ground Truth of Accuracy\n",
+    "- (Optional) Incorrectly Classified Image Files will be copied to a seprate folder with detected class label\n",
+    "- Ground Truth Classification Confusion Matrix is Created\n",
+    "    - (Optional) Confusion Matrix can be Exported as a '.png' file\n",
+    "\n",
+    "---"
    ]
   },
   {
@@ -216,10 +223,18 @@
     "plt.show()"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "98f5ddd2-ee51-4587-bdcc-8e1fba479f43",
+   "metadata": {},
+   "source": [
+    "More At : https://github.com/iSiddharth20/DeepLearning-ImageClassification-Toolkit"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "50710aca-1c65-4477-8387-0815faf93959",
+   "id": "0ddd3728-6672-4a7c-a3b0-e963ee8d119d",
    "metadata": {},
    "outputs": [],
    "source": []

notebooks/deep_learning/CNNImageClassification/HelperFunctions.py

@@ -1,4 +1,9 @@
 '''
+
+Understanding HelperFunctions
+    - Function to display 2 images side-by-side on screen
+    - Function to Extract largest object from souruce image
+
 Source : https://github.com/iSiddharth20/DeepLearning-ImageClassification-Toolkit
 '''